Intergroup bias and self-evaluation: domain-specific self-esteem, threats to identity and dimensional importance

Recent theoretical statements argue that two predictions about self-esteem and intergroup bias may be derived from social identity theory. The first suggests that in-group bias enhances self-esteem. The second suggests that threats to self-esteem enhance in-group bias. Previous research has provided little firm evidence for either prediction. Much of this work has, however, tended to ignore realistic groups, domain-specific self-esteem, the assessment of self-esteem prior to and following the display of in-group bias and the importance attached to evaluative dimensions. The current investigation sought to redress these issues. In separate experiments, men and women presented with positive and negative feedback, evaluated in-group and out-group success and failure on dimensions important and unimportant to the in-group. Women displayed out-group bias on dimensions unimportant to their in-group (i.e. physical ability) with respect to evaluations of failure. Women subsequently manifested a decrease in physical self-esteem. Men displayed in-group bias on dimensions unimportant to their in-group (i.e. verbal ability) and, following positive feedback, on dimensions important to the in-group (i.e. physical ability). Men subsequently manifested an increase in verbal and physical self-esteem. Global self-esteem was unaffected in each study. These findings suggest that the relative display of bias can affect domain-specific self-esteem. No support was found for the postulate that low or threatened self-esteem can enhance bias on evaluative dimensions important to the in-group.